The article doesn't mention it explicitly, but this is a nice example of how using Bayes theorem helps you ignore the hard-to-compute normalization term of the input space. In the article, this is the P(w) term of
P(c|w) = P(c)P(w|c)/P(w),
where c is a correction, and w is the original word.
The author does implicitly talk about this when he explains that P(c|w) conflates the two factors, but it's also not that hard to see that getting a handle on P(w) -- the probability space of misspellings -- is harder than getting a hold of P(c) -- the probability space of actual words, and Bayes lets us get rid of the former during optimization.
Selenium is just a tool built on top of the WebDriver API. One of its main disadvantages is needing to run a complicated proxy program (like geckodriver, ChromeDriver, etc.) built individually for each browser in order to drive your instance. As a result, users sometimes suffer from hard to debug edge cases and other pain points.
They also make interacting with JavaScript on the page a bit painful. For example, injecting JavaScript into the browser with Selenium can be quite an ordeal [1], so you're somewhat limited in what you can do by what Selenium's developers decided to focus on. It also complicates deployments by adding another moving part to the overall equation.
In contrast, the Web Extension API is now part of all major browsers, and makes interacting with different page contexts effortless. To give a sense of the project, we wrote an interactive tour of Remote Browser which runs browser instances on our backend.
You're writing that users suffer from edge cases from the individual webdriver implementations. My experience as someone working with gecko and chromedriver on a daily basis is that the number of edge cases stemming from browser behavior (such as moving to, clicking, and focusing on elements) is a much more frequent pain than differences in the webdriver implementation.
For the benefits of others: spent quite a bit of time googling that exodus thing (way too ubiquitous), and I think this is it: https://github.com/intoli/exodus
If you are interested in Arch but not the overhead it entails, you may wish to consider Antergos https://antergos.com/
It's basically a fully set up Arch with sane defaults. Lets you have the rolling updates and other benefits, without requiring as much energy to set up.
Ubuntu would crash all the time and just was so bloated/ugly and crappy. Been on Antergos almost 8 months now-and it is so awesome, totally love it, runs way smoother.
I use i3+gnome configuration, if you've never used i3 I highly recommend giving it a shot.
Almost no update problems in a few years (although the problems I have had where really bad). I'd recommend it though as I have had more issues with Ubuntu updates and managing a million Papa's.
I actually touch on the relation to whitening toward the bottom of the article. You can whiten your dataset from the left singular matrix U which is directly related to PCs. Thanks for reading!
Great article! I'm the product manager for the headless feature, and I'm unsure why `binary.add_command_line_options('-headless')` doesn't work for you, as the flag does work when invoking Firefox directly:
Thanks! That was guess as well, since running the command you posted from cmd does work. I dug around selenium's source code a bit, but decided to take the pragmatic path and use the MOZ_HEADLESS environment variable instead.
It's funny that New Vegas stands out from the lot; turns out it had some of the original Interplay/Black Isle people working on it. I learned this from a Tyranny AMA on Reddit...
Is anybody else deeply disappointed by the camera bulge? If you use the tablet on a tabletop, that basically forces you to buy a ~$100 case that adds unnecessary thickness and weight to the tablet. Cheap move, IMO.
Seriously. Wtf. I place my tablet down flat on my counter tops 24/7.
I dont even use my tablet camera!! How could they justify making the device uneven for a camera few people use, and for a camera where the old one worked perfectly fine for its purpose.
Seriously wtf move by apple. The camera bump a game-ruiner. I'm irrationally pretty upset by this.
Yes. In reviews of the original iPad Pro most people say that it is a tool best used for laying flat and using the pen. You can't do that with a bulge without a case (Back+Front=$118?!)
But they can be inconvenient. I have a cover on my air 2 to protect the glass but not on the back. It slides more easily into and out of my bag this way.
My 6s+ has the identical lens protrusion -- much less than the bulge on the 6 -- less than a mm. It feels like it lies flat on the table -- certainly it doesn't wobble. The iPad, with its greater baseline, will have even less reason to wobble.
I first heard this performed on a podcast called, "The Truth: Movies for Your Ears." Very good podcast for dramatic stories, sometimes artsy like this one.
Great podcast certainly, and I found it much more enjoyable than the short film adaptation mentioned higher in this thread.
I also liked the story on The Truth about the beings who discovered the Voyager 1 record (http://thetruthpodcast.com/Story/Entries/2014/5/14_Voyager_F...). Although the 'aliens' in that story were far too human and contemporary to take seriously, it worked as a great examination of how our civilization would react to a similar discovery. Although, like many Truth episodes, the light improv style works only well enough for it to be enjoyable. I much prefer their scripted and adapted stories.
I have a 4s. While iOS 7 is laggier than iOS 6 when doing things like switching apps using a home key double tap etc., it is a much faster interface. That is, there might be (not too heavy) laginess for a small set of tasks, but iOS 7 unlike iOS 6 can keep up with my natural speed for doing things much better.
I think the key is that iOS 7 is 'slower', but the animations are meant to take less time. Thus swiping back and forth through view hierarchies, changing apps, etc. is all faster because the animations are designed to take less time (and animation delay is an artificially induced latency).
P(c|w) = P(c)P(w|c)/P(w),
where c is a correction, and w is the original word.
The author does implicitly talk about this when he explains that P(c|w) conflates the two factors, but it's also not that hard to see that getting a handle on P(w) -- the probability space of misspellings -- is harder than getting a hold of P(c) -- the probability space of actual words, and Bayes lets us get rid of the former during optimization.